Rare earth element cycling and reaction path modeling across the chemocline of the Pettaquamscutt River estuary, Rhode Island

Abstract

Rare earth elements (REEs) were measured in the water column of the highly stratified and euxinic Upper Basin of the Pettaquamscutt River estuary (Rhode Island) to investigate their cycling and evolution across the chemocline. The euxinic bottom waters of the estuary exhibit total dissolved sulfide (i.e., ΣS^-II) concentrations up to 4 mmol kg⁻¹, which exceeds values of the Black Sea by nearly 10-fold and makes this system ideal for study of redox-sensitive elements. The REE concentrations decrease with depth in the oxic surface waters, before substantially increasing across the chemocline, corresponding to redox cycling of manganese (Mn) and iron (Fe). For example, the mean ± 1σ neodymium (Nd) concentration in the oxic surface waters is 511 ± 184 pmol kg⁻¹, but increases by approximately 5-fold across the chemocline reaching 2476 pmol kg⁻¹ at 7.4 m depth. Concurrently, Mn and Fe concentrations increase by 5- and 11-fold, respectively, within the chemocline relative to their oxic surface water values. The pore water within the Upper Basin sediments exhibits REE concentrations that are on average 130-fold greater than their values in the overlying waters, driving an upward diffusive REE flux into the euxinic bottom waters. An outstanding feature of the REEs in this study is the preservation of negative cerium (Ce) anomalies across the chemocline into the highly euxinic bottom waters. Likewise, our reaction path model shows that the mass balance of the local water sources is the major driver of the REE compositions of the water column. Primarily, the negative Ce anomaly in the euxinic bottom waters highlights the influence of the Ce depleted sediment pore water on the REE composition and fractionation patterns of the euxinic bottom waters. This study underscores the importance of local sources and indicates that negative Ce anomalies can persist in highly euxinic waters that evolve from Ce depleted sources, further demonstrating that negative Ce anomalies are not necessarily a good proxy of the redox conditions of aqueous environments.

Introduction

Anoxic basins provide a natural laboratory for studying the impact of oxidation-reduction (redox) reactions on trace element cycling, including the rare earth elements (REEs, De Baar et al., 1988, Sholkovitz and Elderfield, 1988, German and Elderfield, 1989, Schijf et al., 1995). Except for cerium (Ce) and europium (Eu) that have multiple oxidation states (Ce³⁺, Ce⁴⁺; Eu²⁺, Eu³⁺), the trivalent REEs lack intrinsic redox chemistry but are nonetheless strongly influenced by the redox cycling of Fe and Mn (Byrne and Sholkovitz, 1996, and references therein). Indeed, several studies of REEs in natural waters have reported strong relationships between the REEs and both Fe and Mn, chiefly because these latter two metals form oxides/oxyhydroxides that scavenge dissolved REEs from oxic natural waters and dissolve in anoxic waters, releasing sorbed metals such as the REEs (e.g., Bau et al., 1996, Ohta and Kawabe, 2001, Quinn et al., 2004). Notably, the absolute concentrations, fractionation patterns and anomalous behavior of select REEs in anoxic basins provide useful information that may allow for the use of REEs as tracers of paleo-ocean circulation (e.g., Elderfield et al., 1990, Holser, 1997), paleo-redox conditions (e.g., German and Elderfield, 1990, Feng et al., 2009), and fluid composition during early diagenesis (e.g., Himmler et al., 2010, Zwicker et al., 2018). Therefore, there is a need to improve our understanding of REE fluxes and cycling across strong redox gradients to better comprehend their biogeochemical cycles in the environment.

Owing to the different valences exhibited by Ce and Eu, both of these REEs can be fractionated from their strictly trivalent REEs neighbors as a result of changing redox conditions, commonly referred to as anomalous geochemical behavior (e.g., Byrne and Sholkovitz, 1996). These anomalies are defined by the ratio of the input-normalized concentration of a specific REE measured in the sample to the predicted concentration of the same REE based on interpolation between its nearest neighbors in the REEs series. A positive anomaly is a value greater than 1, whereas a negative anomaly is a value less than 1. The Ce anomaly has long been proposed as a potential tracer of large scale marine anoxia in ancient oceans (e.g., Holland, 1984), although substantial differences in Ce profiles reported from various studies suggest a limitation to this application (e.g., De Baar et al., 1983, German and Elderfield, 1990, Holser, 1997, Feng et al., 2009). Typically, seawater exhibits substantial negative Ce anomalies, owing to the oxidation of Ce³⁺ to the less soluble Ce⁴⁺, which is scavenged from solution by precipitating materials, e.g., Fe and Mn oxides/oxyhydroxides (Goldstein and Jacobsen, 1988, Elderfield et al., 1990, German et al., 1990, Moffett, 1994a, Moffett, 1994b, Klinkhammer et al., 1994). Similarly, hydrothermal effluents from sources such as mid-ocean ridges, exhibit positive Eu anomalies that are reflective of seawater-rock interactions involving basalts that contain Eu-enriched plagioclase (German et al., 1990, Klinkhammer et al., 1994).

Although the stratified basin waters and sediment pore waters are biogeochemically different, the redox reactions that occur in the anoxic bottom waters of stratified basins may provide some insights into reactions generated during early diagenesis within sediment pore waters (e.g., Elderfield and Sholkovitz, 1987, De Baar et al., 1988, O’Sullivan et al., 1997). Therefore, the relative ease of sampling aquatic anoxic systems compared to sediment core collection and pore water extraction provides the opportunity to study processes that occur during sediment diagenesis and the sedimentary record by sampling the chemocline in stratified basins (O’Sullivan et al., 1997). A better understanding of the cycling of REEs, particularly Ce, in anoxic water columns should enhance our understanding of REE behavior during early diagenesis in sediment pore waters, and possibly provide additional insight into the formation and preservation of Ce anomalies in the marine sedimentary record (e.g., Haley et al., 2004). Processes that may be better understood from this analogy also include diagenetic fluxes of REEs from sediment pore waters; reductive mobilization of REEs from anoxic shelf sediments; and REE enrichment in metalliferous sediments and phosphorites during early diagenesis (De Baar et al., 1988, Murray et al., 1990, Haley et al., 2004, Himmler et al., 2010, Abbot et al., 2015).

Investigations of REE cycling across the chemocline in euxinic basins have included the Black Sea, Saanich Inlet, Famvaren Fjord, and the Cariaco Trench (e.g., Anderson et al., 1988, De Baar et al., 1988, German and Elderfield, 1989, German et al., 1991, Schijf et al., 1994). These studies report increases in dissolved REE concentrations across the chemocline; the coupling of the strictly trivalent REEs to the active redox cycles of Fe and Mn; and preferential release of Ce to the anoxic bottom waters as particulate Mn⁴⁺ and Fe³⁺ are reduced, coupled with Ce⁴⁺ reduction to Ce³⁺ (De Baar et al., 1988, Sholkovitz and Elderfield, 1988, German and Elderfield, 1989, German and Elderfield, 1990, German et al., 1991, Schijf et al., 1995). More specifically, the REEs are scavenged from the oxic surface water by precipitating Fe/Mn oxides/oxyhydroxides and subsequently released back into solution within the chemocline upon reductive dissolution of the settling metal oxides/oxyhydroxides. Although the Black Sea and Cariaco Trench are permanently stratified and euxinic (i.e., mean ± 1σ concentration of H₂S_(aq) below 800 m depth in the Black Sea is 0.044 ± 0.06 mmol kg⁻¹ and is 0.038 ± 0.015 mmol kg⁻¹ below 375 m depth in the Cariaco Trench; Luther et al., 1991, Zhang and Millero, 1993, Dellwig et al., 2019), to the best of our knowledge, the REE geochemistry of more sulfidic waters such as those of Framvaren Fjord, has not been investigated. Dissolved sulfide concentrations in the bottom waters of Framvaren Fjord are approximately 17 times higher than in the Black Sea, exhibiting a mean ± 1σ H₂S_(aq) concentration of 1.52 ± 0.97 mmol kg⁻¹, with ΣS^-II as high as approximately 8 mmol kg⁻¹ (Anderson et al., 1988, Haraldsson and Westerlund, 1988, Luther et al., 1991, Yao and Millero, 1995, Dellwig et al., 2019). The effects, if any, of these highly euxinic waters on REE concentrations and fractionation patterns remain uninvestigated.

Here, we present new REE concentrations data as a function of depth in the Upper Basin of the Pettaquamscutt River estuary, a highly stratified estuary with semi-permanently (i.e., subject to rare overturn events; Orr and Gaines, 1973) euxinic bottom waters where ΣS^-II concentrations up to approximately 4 mmol kg⁻¹ have been reported (e.g., Gaines and Pilson, 1972, O’Sullivan et al., 1997, Wilkin and Barnes, 1997). We build on our earlier investigation that highlighted the importance of submarine groundwater discharge (SGD), which contributes 25% to the REE budget and compositions of the overlying surface waters in the Pettaquamscutt River estuary (Chevis et al., 2015). Specifically, we combine field sampling and measurements of the REE concentrations in discrete water samples and sediment pore water with geochemical modeling to understand the origin and cycling of the REEs in the system. Additionally, we compare our observations of the REE cycling in the Upper Basin of the Pettaquamscutt River estuary to other major anoxic basins.